Aberrant gene expression in affected tissue as compared to normal tissue is a common characteristic of many human diseases. This is true for cancer and many neurological diseases which are characterized by changes in gene expression patterns. Gene expression patterns are controlled at multiple levels in the cell. Control of gene expression can occur through modifications of DNA: DNA promoter methylation is associated with suppression of gene expression. Several inhibitors of DNA methylation are approved for clinical use including the blockbuster Vidaza™. Another class of modifications involve histones which form the protein scaffold that DNA is normally associated with (coiled around) in cukaryotic cells. Histones play a crucial role in organizing DNA and the regulated coiling and uncoiling of DNA around the histones is critical in controlling gene expression—coiled DNA is typically not accessible for gene transcription. A number of bistone modification have been discovered including histone acetylation, histone lysine methylation, histone arginine methylation, histone ubiquinylation, and histone sumoylation, many of which modify accessibility to the associated DNA by the cells transcriptional machinery. These histone marks serve to recruit various protein complexes involved in transcription and repression. An increasing number of studies are painting an intricate picture of how various combinations of histone marks control gene expression in cell-type specific manner and a new term has been coined to capture this concept: the histone code.
The prototypical histone mark is histone acetylation. Histone acetyl transferase and histone deacetylases are the catalytic machines involved in modulation of this histone mark although typically these enzymes are parts of multiprotein complexes containing other proteins involved in reading and modifying histone marks. The components of these protein complexes are typically cell type and typically comprise transcriptional regulators, repressors, co-repressors, receptors associated with gene expression modulation (e.g., estrogen or androgen receptor). Histone deacetylase inhibitors alter the histone acetylation profile of chromatin. Accordingly, histone deacetylase inhibitors like SAHA, TSA, and many others have been shown to alter gene expression in various in vitro and in vivo animal models. Clinically, histone deacetylase inhibitor have demonstrated activity in the cancer setting and are being investigated for oncology indications as well as for neurological conditions and other diseases.
Another modification that is involved in regulating gene expression is histone methylation including lysine and arginine methylation. The methylation status of histone lysines has recently been shown to be important in dynamically regulating gene expression.
A group of enzymes known as histone lysine methyl transferases and histone lysine demethylases are involved in histone lysine modifications. One particular human histone lysine demethylase enzyme called Lysine Specific Demethylase-1 (LSD1) was recently discovered (Shi et al. (2004) Cell 119:941) to be involved in this crucial histone modification. LSD1 has a fair degree of structural similarity, and amino acid identity/homology to polyamine oxidases and monoamine oxidases, all of which (i.e., MAO-A, MAO-B and LSD1) are flavin dependent amine oxidases which catalyze the oxidation of nitrogen-hydrogen bonds and/or nitrogen carbon bonds.
Several groups have reported LSD1 inhibitors in the literature. Sharma et al. recently reported a new series of urea and thiourea analogs based on an earlier series of polyamines which were shown to inhibit LSD1 and modulate histone methylation and gene expression in cells (J. Med. Chem. 2010 PMID: 20568780). Sharma et al. note that “To date, only a few existing compounds have been shown to inhibit LSD1.” Some efforts were made to make analogs of the histone peptide that is methylated by the enzyme, other efforts have focused on more small molecule like molecules based on known MAO inhibitors. Gooden et al. reported trans-2-arylcyclopropylamine analogues that inhibit LSD1 with Ki values in the range of 188-566 micromolar (Gooden et al. ((2008) Bioorg. Med. Chem. Let. 18:3047-3051)). Most of these compounds were more potent against MAO-A as compared to MAO-B. Ueda et al. ((2009) J. Am. Chem. Soc. 131(48):17536-17537) reported cyclopropylamine analogs selective for LSD1 over MAO-A and MAO-B that were designed based on reported X-ray crystal structures of these enzymes with a phenylcyclopropylamine-FAD adduct and a FAD-N-propargyl lysine peptide. The reported IC50 value for phenylcyclopropylamine was about 32 micromolar for LSD1 whereas compounds 1 and 2 had values of 2.5 and 1.9 micromolar respectively.
Binda et al. examined a series of phenylcyclopropylamine derivatives in relation to their inhibitory activity against LSD1 and LSD2 as well as examining stereochemical issues in relation to the cyclopropyl ring (J. Am. Chem. Soc. 2010 May 19; 132(19):6827-33). Binda et al. reported that their para substituted phenylcyclopropylamine derivatives are non-selective which as a group are appear to be better MAO-A inhibitors than MAO-B inhibitors. Furthermore, their inhibitory activities against MAO-A and LSD1 were roughly the same.
Mimasu et al. disclose a series of phenylcyclopropylamine derivatives having benzoyl substitutions at the ortho-position ((2010) Biochemistry PMID: 20568732. Ortho-substituted compounds from this series without a benzoyl group in the ortho-position e.g., phenyl, alkoxy, or having a combination of ortho- and para-substitution appeared to be less potent inhibitors of LSD1 than those compounds having benzoyl substituents in the ortho-position. The most active compounds from this series had a benzoyl group at the ortho-position and one or two meta-fluoro substitutions: biphenyls like S1310 and compounds having large groups in the para-position were less effective LSD1 inhibitors.
The phenylcyclopropylamines have been the subject of many studies designed to elucidate a SAR for MAO inhibition. Kaiser et al. ((1962) J. Med. Chem. 5:1243-1265); Zirklc et al. ((1962) J. Med. Chem. 1265-1284; U.S. Pat. Nos. 3,365,458; 3,471,522; 3,532,749) have disclosed the synthesis and activity of a number of phenylcyclopropylamine related compounds. Other phenylcyclopropylamine type compounds are disclosed in Bolesov et al. ((1974) Zhurnal Organicheskoi Khimii 10:8 1661-1669) and Russian Patent No. 230169 (19681030).
Studies have been conducted with phenylcyclopropylamine related compounds to determine selectivity for MAO-A versus MAO-B since MAO-A inhibitors can cause dangerous side-effects (see e.g., Yoshida et al. (2004) Bioorg. Med. Chem. 12(10):2645-2652; Hruschka et al. (2008) Biorg Med Chem. (16):7148-7166; Folks et al. (1983) J. Clin. Psychopharmacol. (3)249; and Youdim et al. (1983) Mod. Probl. Pharmacopsychiatry (19):63).